Superconducting Electronics for Metrology...New proposed SI Unit Actual SI NEW SI second, s...

Superconducting Electronics

for Metrology

Paul Dresselhaus

Quantum Voltage Project

Superconducting Electronics Group

National Institute of Standards and Technology

Boulder, CO, USA

Sam Benz, Alain Rufenacht, Nathan Flowers-Jacobs, Horst Rogalla, Anna Fox

June 16, 2017

IEEE/CSC & ESAS SUPERCONDUCTIVITY NEWS FORUM (global edition), No. 41, July 2017. This keynote presentation Fr-KEY-01 was given at ISEC 2017.

1

Applications of JVS systems

10 V PJVS

2 V JAWS

2 µV QVNS

“Electronic”

kilogram

Planck

constant

Electrical

power

standard

(diff.sampl.)DVM gain &

linearity, ratio

calibration

Zener

reference

calibration

AC voltmeter

calibration

Therm. trans.

standard

calibration

Impedance

ratio

calibration

“Electronic”

thermometry

Boltzmann

constant

JVS

SYSTEMS


2

New proposed SI

Unit Actual SI NEW SI

second, s Dn(133Cs)hfs Dn(133Cs)hfs Cs hyperfine transition

meter, m c c Speed of light

kilogram, kg m(K) h Planck constant

ampere, A µ0 e Elementary charge

kelvin, K TTPW k Boltzmann constant

mole, mol M(12C) NA Avogadro constant

candela cd Kcd KcdLight source intensity at a frequency of 540 THz


3

Features of a Quantum Voltage Standard

• Intrinsic accuracy is based on quantum behavior

– Josephson effect defines the electrical properties of

superconducting Josephson junctions

• Always produces an accurate voltage

– Regardless of environmental conditions or location, which

is in contrast to “artifact” standards, and

– Over a range of all bias parameters and operating

conditions, called “flat spots”.

• Voltage is only correct on-chip

– Systematic errors must be removed or characterized

– Measurement leads & circuit parasitics (“leakage” R, L

and C) have thermal voltages & error currents


4

Single

Josephson

Junctions

Single

Josephson

Junctions

Artifact Standards for DC Voltage

Varies in time & with

environmental

conditions

10 volt

Josephson

array chip

Weston Cell,

electrochemical

battery cell

Intrinsically accurate

based on quantum

behavior of Josephson

junctions

1V

Weston Cells

10 mV SingleJunctions

1-10V

Arrays

Replaced by Josephson Voltage Standards


5

• Phase quantization ensures quantized voltage pulses

• One voltage pulse for every 2 phase change

• Pulse area is always EXACTLY one flux quantum

Quantum Behavior of Josephson Junctions

dt

d

e

htV

2

1

2

e

hdt

e

hdttV

22

1

2)(

2

00

VGHzKe

h

J /4835979.0

11

20


6

Driven Damped Josephson Junction

• DC current bias, I

• Continuous microwave or periodic pulse bias

– Supercurrent oscillations entrain to the drive frequency f

– Lock at harmonic integers n

– Generate constant DC voltage steps Vn

– Over a dc bias current range or flat spot

fe

hnVn

2

V

I

n=+1

n=0

n=1

V

I

R

Ic

Bias

I V


7

Practical Voltages Require Series Arrays

• Single junction voltages 20 µV

• 10 V is desired output voltage

• Large series arrays are required

• Uniform junctions

• Uniform microwave power

• Frequency accuracy derived from atomic

clock

nNfe

hVn

2

N Accurate

Voltage

Practical Voltages Require Long, Uniform Series Arrays

GHzVe

h2

2

V

I

No Common

Bias


8

Programmable Josephson Voltage Standard


9

Programmable Josephson Voltage Standard

• Sub-array JJs in ternary sequence N=1, 3, 9, ….

• 20 GHz microwave bias to all sub-arrays

• DC current bias to each sub-array selects

one of three voltage steps, n

fe

hnNVArray

2

Hamilton, Burroughs and Kautz, 1995

V

I

n = +1

n = 1

–IBias

+IBias

n = 0

20 GHz

Bias 1 3 N = 9

=4

DC Current Biases

Output Voltage

Filters


10

Fabrication & Design of Superconducting Circuits

• Boulder Micro-

Fabrication

Facility

• Superconducting integrated circuits

– Uniform junctions, barrier materials,

low-defect fabrication

• Microwave circuit design

– Lumped element inductors &

capacitors, power splitters, coplanar

waveguides, simulation & modelling

32 Arrays

with

265,116 JJs

DC Input/Output Pads

(12 x 17) mm2 PJVS Chip

Microwave Input

Splitters


11

Current (mA)

Arr

ay V

olt

age

(V) 2 mA Flat Step Width

Variation of BOTH junctions

and microwave power in arrays

Flat Spots of 16 Arrays with 16800 Junctions

Uniform junctions and microwaves

18 GHz


12

• Optimized microwave, thermal and cryogenic design

• Interchangeable operation in liquid helium probes or cryostat

• Connectorization enables solder-free, fast mounting

• Reconfigurable sub-array with flex matrix bias leads

Top Bottom

PJVS Cryopackage


13

NIST Cryocooled PJVS System

• Integrated system

– Bias electronics DC & MW

– Cryogenics

– Superconducting devices

– Turn-key integrated system

– Automation software

• Optimize & check quantum

states, flat spots

• Performs measurements

• Specific measurement

techniques needed for

different applications

Compressor

Cryostat &

MW Amplifier

4 K Cold Plate


14

Leakage currents in the PJVS system

• Need high Leakage impedance

(CJB& ALR)

– PJVS (dc leakage resistance)

• 10 V/10 G Ω = 1nA

• 1 nA * 1 Ω = 1 nV– 1 part in 1010 at 10 V

Leakage resistance should be ~> 100 G Ω


15

PJVS Applications & Best Results

• Stable DC & stepwise AC voltages

– Calibrate:

• Zener reference standards

• DVM gain, linearity

• Calibrators

• Power meters

– Precision Measurements:

• Electronic kilogram for measuring Planck’s constant, h

• Uncertainty for different measurement techniques

– DC to DC comparison of PJVS: 3 x 10-11

– Differential sampling frequencies < 500 Hz: < 1 µV/V


16

Intrinsically Accurate AC Synthesis


17

AC Voltage Standards &

Arbitrary waveform Synthesizers• Step-wise approximated sine waves

– Transitions between steps compromise accuracy

• Direct digital synthesis with current pulse sequences

– Intrinsically accurate by controlling every quantized pulse

V(t)

t

( )2

pulse

hV t

e

V(t)

t

Transitions

Steps


18

Pulse-Driven Voltage Synthesis

• Arbitrary voltage waveforms

• Pulse biased

• Directly control every JJ pulse– Digital signal provided by high-speed

pulse generator

• Waveform synthesis– Timing and polarity precisely

determine the voltage waveform

• Intrinsically accurate – Pulse density determines voltage

– JJ pulses are perfectly quantized

Co-invented in 1995 by NIST & Westinghouse researchers,

A.H. Worsham, J.X. Przybysz, S. Benz, and C. Hamilton

–√2V

Time

√2V

1 ms

1 kHz sine wave

Control 3 quantum states (+, 0, pulses) for each junction

to produce bipolar waveforms


19

“Artifact” Detectors vs. “Quantum” Sources

• Conventional AC standards are RMS detectors

– Thermally AC DC voltage signals

– “Artifact” standards have similar performance, but NOT identical

Transfer Standards

Thermal Converters

• Provide a source based on quantum effects


20

First Comparisons of 1 V AC

Josephson Voltage Standard Sources

• Statistical uncertainty of first intercomparisons of 1 V AC JAWS

voltages are now below 0.1 V/V

• Systematic errors are ~1 V/V for kilohertz frequencies

1 V

JAWS

Commercial

Voltage

Calibrators

Weston Cells

SingleJunctions

PJVS

Calibrators

JAWS

DC AC


21

1V rms JAWS Chip and Circuit

• 51,240 total junctions in 4 arrays

• 2 pulse biases and 4 low-speed biases

• Clocking pulses at 15 GHz = 770x1012 quantum states/s

• Measure spectra with a 24-bit ADC digitizer

1 cm x 1 cm Chip


22

2016: 2 V cryocooled system,

4 pulse channels (2 sets of bias electronics)

Advances in the JAWS system

1 V ± 1 V output

Pulse2Pulse2Pulse1Pulse1

4.3 K cryocooler


23

Flat Spot of Quantum States

Nathan Flowers-Jacobs, 2016

1.1 mA DC

Bias Current

Difference from Sine Voltage

V(Measured) - V(Calc Sine)


24

-165 dBc from 2 V

92 nV rms

(nominally 0 V

rms)

1 V ± 1 V JAWS comparison at 1 kHz• Null measurement

2.0000

V rms

nominally

0 V rms

quadrature (V rms)

in-p

has

e (V

rm

s)

1

2

00 1-1

20 nV rms

Insensitive toADC

gain/linearity

Sensitive tosingle-pulsetime delays


25

JAWS Applications

• Calibrations:

– Voltage calibrators

– AC-DC transfer standards

– Thermal converters

• May require trans-impedance amplifier

• Characterizing analog and digital electronics:

– Gain, linearity, and distortion of analog-to-digital converters

(ADCs) and amplifiers

• Impedance comparisons (*New)


26

Josephson-Based Full Digital Bridge for high-

accuracy impedance comparisons

• Voltage ratio can be set to any desired value and frequency

– Don’t need different transformers to match different impedance ratios

– Allows direct comparison between the decadal scale and AC-QHE, 10:12.906

• Any impedance in the complex plane can be calibrated

– Because the phase between the JAWS sources can be adjusted to any value

Overney et al., Metrologia, June 2016


27

Quantum Voltage Noise Source


28

Artifact Standards in

Thermometry

• ITS-90 scale of fixed points

• K defined by triple point of water

• Scale is designed to “represent”

thermodynamic temperature

• Interpolation required for temps

between fixed points


29

(52 nV)2 rms

1600 Hz harmonics

Johnson Noise Thermometry with a

Quantized Voltage Noise Source

• Fully electrical measurement of thermodynamic temperature

• Measure Johnson noise voltage of a 100 Ohm resistor at different temperatures

• SMALL 2 nV/√Hz noise voltage amplifier noise voltage

• Requires cross-correlation measurement techniques

• Calibrate electronics with quantum-accurate “pseudo-noise” voltage synthesized with a QVNS

2 4 D DT TV S f kTR f


30

Johnson Noise Thermometry

• Compare thermal and electrical noise powers

• Temperature is determined by noise ratio, measured

resistance R, fundamental constants h, k and Rk, and

QVNS constants D, N, fclock and M)

T

Q

S

S

Water Triple Point

R, Johnson Noise

Quantized Voltage

Noise Source (QVNS)

Noise

Voltage

Correlator

SQ

ST

16

22 MNDf

R

R

S

S

k

hT clock

k

Q

T


31

Quantum Voltage Noise Source QVNS

8 junctions

Cryopackage

1 cm2

Chip


32

NIST Johnson noise thermometer

Resistor,

Triple Point Cell, and

Thermoelectric Cooler

Chargers

Batteries

Cross-Correlation

Electronics

QVNS in LHe Dewar

Pulse Bias Source is

outside shielded room


33

Johnson Noise Thermometry Results

• Electronic determination of Boltzmann’s constant, k

– Uncertainty results

• 1210-6 (5ppm Type A in 10 days) NIST USA, 2010

• 410-6 (3ppm Type A in 34 days) NIM China, 2015

• (9ppm Type A in 7 days) AIST Japan, 2016

– < 310-6 is the goal for SI redefinition

• Primary “thermodynamic” thermometer

– Measured Sn, Ga and water on ITS-90 temperature scale

– Perfect voltage and temperature linearity has potential to

increase knowledge of ITS-90 fixed-point “artifacts” and

eliminate interpolation between temps

16

22 MND

T

f

R

R

S

S

h

k

TPW

clockk

Q

T


34

Summary• Programmable Josephson Voltage Standard (PJVS)

– Intrinsically accurate DC voltages

– ±10V peak stepwise-approximated AC waveforms

– Requires differential sampling on steps to realize quantum accuracy

– Applications: DC voltage, AC voltage < 200 Hz, AC power

• Josephson Arbitrary Waveform Synthesizer (JAWS)

– Intrinsically accurate AC waveforms

– 2 V RMS sine waves, or 5.6 V peak-to-peak arbitrary waveforms

– Challenge is sourcing current to low-impedance meters

– Applications: DC voltage, AC voltage, AC power, Impedance

• Pulse-driven Quantum Voltage Noise Source (QVNS)

– Intrinsically accurate arbitrary waveforms

– 2 V peak arbitrary waveforms of harmonic combs

– Applications: Johnson noise thermometry, primary thermometry

Microwave

Bias

Pulse

Bias


35

Future of JVS for SCE

• Quantum voltage standards continue to:

– Improve in performance (V, margins, frequency)

– Replace artifact standards (DC, AC, temperature)

– Be applied in precision measurement that require quantum

accurate performance (kg, h, k, impedance, RF metrology

and perhaps quantum computing).


36

Superconductive Electronics Group in Boulder, Colorado

Thank you!


37

Flat Spots at 10 V

Alain Rufenacht, 2016


38

Flat Spots at 10 V

Alain Rufenacht, 2016


39

Flat Spots at 0 V (+5 V and 5 V)


40

Systems & Circuits

60 Conventional JVS (incl. Hypres, Inc.)

18 Programmable JVS

8 Josephson Arbitrary Waveform

Synthesizer

4 Johnson Noise Thermometer

Measurement Labs

National Measurement Institutes: international

National Laboratories: NASA, Sandia

Industry: HP, Lockheed, Fluke, Keithley, Boeing

DOD: Army, Navy, Air Force

DOE: Oak Ridge Nat. Lab.

NIST Voltage Standard Systems and Circuits


41

Uncertainty of RMS Detector-based Calibrations

1000 V

100 V

10 V

1 V

100 mV

10 mV

1 mV

10 Hz

100 Hz

1 kHz

10 kHz

100 kHz

1 MHz

5

1 V/V

10

20

100

300

50200100

1000

Frequency

Vo

ltag

e

Original JAWS

JAWS Region

of Impact

10x-100x

reduction

1 Hz

New 2V JAWS

Tom Lipe, 2005


42

Superconducting Electronics for Metrology...New proposed SI Unit Actual SI NEW SI second, s...

Documents

Transcript of Superconducting Electronics for Metrology...New proposed SI Unit Actual SI NEW SI second, s...